Developing coarse-grained models for agglomerate growth

In this paper we present a coarse-graining (CG) approach for the agglomeration of nano-particles and clusters. In the current context, coarse-graining involves the replacement of fractal-like clusters by "representative" spherical particles. This simplification reduces significantly the number of degrees of freedom and allows for the computation of much larger systems and for better collision statistics of larger clusters. However, detailed information on the cluster shape is lost, but it is exactly this detailed shape that determines collision frequencies between fractal clusters and thus the agglomerates' growth. Therefore, additional properties need to be "inherited" by the coarsegrained particle that ensure similar collision frequencies. We generate collision probabilities as functions of the minimum passing distance between the clusters and provide these as additional function to the CG particle. This allows for partial overlap between CG particles, and the collision/sticking event is triggered with a specific probability only. We compare collision frequencies of CG simulations with equivalent Langevin dynamics simulations where all primary particles are tracked, and we observe decent agreement between cluster growth predicted by CG and the detailed Langevin dynamics simulations. Remaining differences may stem from differences in cluster transport.